How a new recommendations engine led to gold with a 30% lift in conversion rates

In this Olympic year, I’ve been pondering the striking parallels between Olympic success and the vital role of pay TV operators in delivering content their subscribers desire (no judgment, please). Much like athletes’ dedication to excellence in a fiercely competitive environment, pay TV operators must provide uncompromised, tailored content in their own competitive environment. 

Challenges faced by both athletes and operators mirror hurdles, and the pursuit of cutting-edge solutions aligns with the quest for a competitive edge. Notably, the transformative impact of content recommendations on conversion rates mirrors the triumphant moments of Olympic achievements, highlighting a shared commitment to excellence and surpassing industry benchmarks. 

• How does your UX and metadata compare with your competitors?
• Unlock the Power of Metadata Enrichment

 This was recently shown to be true in the launch of a new recommendation’s engine at a very large pay TV operator.  Although the operator thought they had the engine to win Gold, they began to realize they were on the verge of not medalling at all.  Unfortunately, they were falling short in key areas of success: 

• Retaining valuable subscribers impacting profitability 
• Loosing efficiency costing valuable man hours 

A key issue in this predicament was their recommendations engine wasn’t performing well. What they had, despite its market-leading status, was an engine that had two giant hurdles.   

• It relied solely on perfect metadata sets for content filtering which led to subscribers missing relevant recommendations. 
• It suffered performance bottlenecks during peak times and couldn’t handle concurrent user requests compromising the user experience, causing errors, and requiring significant man-hours for management. 

What they needed for Gold, was an innovative engine which employed a multi-model approach leveraging 2 different filters, content and collaborative, and layered trend analysis on top to deliver robust recommendations across varying quality metadata, even datasets that are not perfect. Additionally, they needed a better way to handle concurrent requests without latency issues.  

• Want a free audit of your metadata and UX? Click here and we’ll help you out!

So, they embarked on a subscriber pilot program with a new engine yielding significant improvements in both the quality of recommendations and the efficiency of the platform. The overwhelmingly positive results led to a full-scale rollout, replacing the underperforming third-party solution. The results speak for themselves. 

• A 30% conversion increase; playback attributed to the recommendations engine  
• A reduction to nearly negligible daily errors; previously 700,000 instances of high-latency responses per day  

These enhancements markedly elevated the end-user experience, setting a new benchmark for recommendation system performance and delivered Gold for the operator. If you seek to improve the overall user experience and provide the content your subscribers desire, explore the solution that achieved Gold synamedia.com/go-recommend/ or set up an exploratory meeting.